Serious bug identified in SmartDashboard/NetworkTables -- robot hangs

[NotInControl]

Hello,

I am joing this converstation late. We just competed in our first district event this past weekend and notices some strange anomolies which maybe contributed to the problem identified here.

We use smartdashboard to send booleans back to a dashboard to let the drivers know certain events or robot states have been reached.

A few times during pit testing we noticed that we could not successfully command the robot. The robot was in a hung state. Comms were up, but the dashboard was frozen, and while a button press did register on the default dash, no response was displayed by the robot. Restarting the Robot, AND restarting the driverstation seemed to be the only way to get arround this and re-establish comms.

I was just writing to make sure I understand the bug because we use Java and it was unclear if this problem was just for C++, on the client side/ ther server side, both...

Based on what I read it appears the crux of the problem is that although NT is multithreaded, it holds on to a lock during a write sequence. A write sequence which also blocks and keeps the lock if the write fails. The robot thread uses this same lock to push data, thus causing the hang on the robot. Is this all correct?

My question is the code to obtain the lock on the robotside in NT or in SmartDash? If you were to call the smartDashboard putXXX methods in a separate thread, would that not at least prevent the hang?

I have not spent any time looking into the NT/SD implemenation, yet, but I will.

Again we are using Java on the robot, and the pyNetorkTables port provided by Dustin on the driverstation.

Regards,
Kevin